What we found
Three diagnostics, unedited.
Names withheld, nothing else. Every figure below was fetched on the date given, by a method we can describe. Where something was found and not yet fixed, it says so — a case study that only lists wins is a brochure.
Dearborn Heights, Michigan · September 2026
A halal bakery
What the audit turned up
The menu page declared its canonical URL as an address that redirected straight back to the page declaring it — a loop invisible to any checker that only confirms a canonical tag exists.
One verified inbound link in total, from a mapping directory. That was the entire profile.
Absent from the halal directory used across the densest Arab-American market in the United States, where twenty-six competitors in the same city were listed.
Three different phone numbers in circulation across her own site, her listings and a stale social profile.
When an answer engine recommended her by name, it was reading a review site — not the website that scored full marks on its own audit.
Fixed, and verified
The canonical loop, at the source rather than page by page, and verified independently afterwards by a second crawler.
The on-site assistant, which had been failing silently: three of three probes returned grounded answers after the fix, where the model behind it had been returning 410 Gone.
Found, not fixed
Her Google Business Profile carries 113 reviews at 4.2 stars and no website link at all. One field, free to change, and it needs her login — not ours.
Montréal, Québec · September 2026
A neighbourhood butcher
What the audit turned up
161 reviews at 4.9 stars, a listed phone number, a real address, and no website of any kind.
Three more like it within the same trade in a single sample of thirty: a fruiterie at 260 reviews, a bakery at 248, a fishmonger at 151 — all operating, all with nothing online to capture what they had already built.
Found, not fixed
Not yet approached. Listed here because the pattern is the finding: demand proven at the owner’s own expense, and none of it captured.
Montréal, Québec · 20 September 2026
A scan of one trade, city-wide
What the audit turned up
Forty-five restaurants sampled from open map data. Eight were permanently closed and seven resolved to a different business entirely — a third of the sample was unusable.
The closed ones were the dangerous half. A shut business keeps its reviews and loses its website link, which is precisely the shape of a perfect prospect. Left unchecked they rank first.
Fixed, and verified
The scan now reads business status and rejects a name that does not match, so a dead storefront cannot present itself as an opportunity.
Answers
About these three.
Why are these anonymous?
Because we have not asked permission to name them, and naming a client without asking is a worse look than withholding a name. Everything except the name is here: the city, the date, the measurement and the method. If you want a named reference before you buy, ask and we will arrange one properly.
Are these results or findings?
Mostly findings, and the page says which is which. Two items were fixed and independently verified. The rest were measured and not yet acted on. A case study that only lists wins is a brochure, and we would rather you could tell the difference.
How were these numbers produced?
By fetching. Review counts and profile fields come from the Google Places API, links and directory presence from direct requests to each source, and the permit figures from the City of Montréal open data release. Nothing here is estimated, projected, or inferred from a sample.
Why is there no traffic or ranking figure anywhere on this page?
Because nothing we run measures those, so any number we printed would be invented. We will tell you a ranking is unavailable rather than guess at one — which is the same standard every report we send you is held to.
The same audit, on your site, free.
A score out of 100 and a ranked list of every fix we find, worst first, each with what it is costing you. Two working days, no call, yours to keep either way.
Start the questionnaire